Spam Email Classifier with Node.js

This project uses Node.js and the Natural library to create an AI-based application that classifies emails as spam or not spam. The application uses a Naive Bayes classifier for spam detection, which is a common algorithm for text classification tasks.

Prerequisites

Before you begin, make sure you have the following installed:

• Node.js: Download Node.js
• npm (Node Package Manager): npm comes with Node.js installation.

Steps to Set Up the Project

Step 1: Set Up Your Project

1. Create a Project Folder: Open your terminal or command prompt and create a new folder for your project.

   mkdir spam-email-classifier
   cd spam-email-classifier

1. Initialize a Node.js Project: Inside the folder, run the following command to create a package.json file.

   npm init -y

Step 2: Install Dependencies

Run the following command to install the required dependencies:

npm install natural

• natural: A library that provides various NLP (Natural Language Processing) tools including classification using Naive Bayes.

Step 3: Create the Spam Classifier

Create a new JavaScript file (e.g., spamClassifier.js) and add the following code:

const natural = require('natural');

// Create a new Naive Bayes classifier
const classifier = new natural.BayesClassifier();

// Sample spam and non-spam data
const spamData = [
  { text: "Congratulations, you've won a 00 gift card!", label: 'spam' },
  { text: "You are eligible for a free trial, click here to sign up.", label: 'spam' },
  { text: "Important meeting tomorrow at 10 AM", label: 'not_spam' },
  { text: "Let's grab lunch this weekend!", label: 'not_spam' }
];

// Add documents to the classifier (training data)
spamData.forEach(item => {
  classifier.addDocument(item.text, item.label);
});

// Train the classifier
classifier.train();

// Function to classify an email
function classifyEmail(emailContent) {
  const result = classifier.classify(emailContent);
  return result === 'spam' ? "This is a spam email" : "This is not a spam email";
}

// Example of using the classifier to detect spam
const testEmail = "Congratulations! You have won a 00 gift card.";
console.log(classifyEmail(testEmail)); // Output: "This is a spam email"

// Save the trained model to a file (optional)
classifier.save('spamClassifier.json', function(err, classifier) {
  if (err) {
    console.log('Error saving classifier:', err);
  } else {
    console.log('Classifier saved successfully!');
  }
});

Step 4: Run the Classifier

To run the classifier, open a terminal and navigate to the project folder. Then, run the following command:

node spamClassifier.js

You should see an output similar to this:

This is a spam email
Classifier saved successfully!

Step 5: Load the Saved Classifier (Optional)

You can load the classifier model later to classify new emails. Here’s how to load the model and classify new emails:

const natural = require('natural');

// Load the saved classifier
natural.BayesClassifier.load('spamClassifier.json', null, function(err, classifier) {
  if (err) {
    console.log('Error loading classifier:', err);
  } else {
    // Classify a new email
    const testEmail = "You have won a free iPhone!";
    console.log(classifier.classify(testEmail)); // Output: 'spam' or 'not_spam'
  }
});

Step 6: Improve the Model (Optional)

To improve the accuracy of the spam classifier, you can:

• Add more training data: Include more samples of spam and non-spam emails.
• Experiment with different algorithms: Try other classification algorithms or models if Naive Bayes is not sufficient for your needs.
• Use advanced techniques: Implement deep learning or neural networks for more complex classification tasks.

Step 7: (Optional) Integrate with Email System

If you want to send or receive emails from the app, you can use the Nodemailer library to send emails.

1. Install Nodemailer:

   mkdir spam-email-classifier
   cd spam-email-classifier

1. Send an Email (Example):

   npm init -y

Conclusion

This guide walked you through setting up an AI app using Node.js and Naive Bayes to classify emails as spam or not spam. You can expand this app by:

• Adding more training data for better accuracy.
• Using more advanced machine learning techniques.
• Integrating the classifier into a web application or email system.

The above is the detailed content of Building a Spam Email Classifier Using AI: A Basic Application. For more information, please follow other related articles on the PHP Chinese website!